Current Status and Potential Applications of Underexplored Prokaryotes

Thousands of prokaryotic genera have been published, but methodological bias in the study of prokaryotes is noted. Prokaryotes that are relatively easy to isolate have been well-studied from multiple aspects. Massive quantities of experimental findings and knowledge generated from the well-known prokaryotic strains are inundating scientific publications. However, researchers may neglect or pay little attention to the uncommon prokaryotes and hard-to-cultivate microorganisms. In this review, we provide a systematic update on the discovery of underexplored culturable and unculturable prokaryotes and discuss the insights accumulated from various research efforts. Examining these neglected prokaryotes may elucidate their novelties and functions and pave the way for their industrial applications. In addition, we hope that this review will prompt the scientific community to reconsider these untapped pragmatic resources.


Introduction
Ample numbers of prokaryote (bacteria and archaea) phyla have been described. The modern classification of prokaryotes involves polyphasic characterizations [1]. The state-of-the-art methods-for instance, phylogenomics, average nucleotide identity, and percentage of conserved proteins-are becoming popular for assisting the delineation of prokaryote genera [2][3][4]. To date (September 2019), more than 3800 prokaryotic genera have been published directly in the International Journal of Systematic and Evolutionary Microbiology (IJSEM) or were included in the validation list, under the Rules of the International Code of Nomenclature of Bacteria [5,6]. The LPSN database (List of Prokaryotic names with Standing in Nomenclature database; http://www.bacterio.net) archives all validly published names of prokaryotes. The BacDive bacterial metadatabase has collected data on more than 80,500 strains, including 13,500 type strains from 34 bacterial and three archaeal phyla [7]. The World Data Centre for Microorganisms (WFCC-MIRCEN; http://www.wdcm.org) is another data center for microbial resources. The WFCC Global Catalogue of Microorganisms (GCM) has catalogued 447,444 strains from 48 countries and regions. Figure 1. Illustration of 16S rRNA phylogenetic tree (phyla). The representatives were aligned using ClustalW, and a maximum-likelihood tree with a bootstrap value of 1000 replicates was built within the MEGA6 package [27]. The transformed radiation tree was redrawn using the FigTree version 1.4.4 program. The number of 16S rRNA sequences for each phylum is provided in Table 1.  Figure 1. Illustration of 16S rRNA phylogenetic tree (phyla). The representatives were aligned using ClustalW, and a maximum-likelihood tree with a bootstrap value of 1000 replicates was built within the MEGA6 package [27]. The transformed radiation tree was redrawn using the FigTree version 1.4.4 program. The number of 16S rRNA sequences for each phylum is provided in Table 1.
In this review, we focus our discussion on the current status and potential applications of underexplored prokaryotes. Herein, we define underexplored prokaryotes as (i) unculturable prokaryotes, and (ii) genera consisting of limited species listed in the LPSN database.

Definition of Underexplored Prokaryotes
The word 'rare' is loosely defined in microbiological, taxonomical, and ecological perspectives. Microbiologists often refer to rare or underexplored prokaryotes as (i) culturable genera with limited type strains, (ii) unculturable microorganisms under laboratory conditions, or (iii) prokaryotes that present a minority population in the environment. The majority of the prokaryotes present in a sample are mostly uncultivable [13,[30][31][32]. These underexplored prokaryotes are often referred to as 'microbial dark matter', or 'unculturable bacteria'. Rare prokaryotes are also microorganisms present as minorities in the environment whose abundance can be as low as <0.001% of the total population [33].

Reasons for Analysing Underexplored Prokaryotes
Below, we give three main motivations behind the interest of research groups working in the field of underexplored prokaryotes. The first motivation is obtaining fundamental knowledge and completing a bigger picture. Trees constructed using currently known cultured taxa are only the 'tip

Definition of Underexplored Prokaryotes
The word 'rare' is loosely defined in microbiological, taxonomical, and ecological perspectives. Microbiologists often refer to rare or underexplored prokaryotes as (i) culturable genera with limited type strains, (ii) unculturable microorganisms under laboratory conditions, or (iii) prokaryotes that present a minority population in the environment. The majority of the prokaryotes present in a sample are mostly uncultivable [13,[30][31][32]. These underexplored prokaryotes are often referred to as 'microbial dark matter', or 'unculturable bacteria'. Rare prokaryotes are also microorganisms present as minorities in the environment whose abundance can be as low as <0.001% of the total population [33].

Reasons for Analysing Underexplored Prokaryotes
Below, we give three main motivations behind the interest of research groups working in the field of underexplored prokaryotes. The first motivation is obtaining fundamental knowledge and completing a bigger picture. Trees constructed using currently known cultured taxa are only the 'tip of the iceberg' of the entire group of existing prokaryotes ( Figure 1). Researchers accept that many prokaryotes are still hidden in the tree of life. Cultivation-independent methods (i.e., 16S rRNA amplicon sequencing, MAG, and SAG) can further expand our understanding of domain bacteria beyond those that have been cultivated [34]. Studying rare prokaryotes generates new scientific information that unveils their uniqueness. Some of these underexplored prokaryotes have atypical characteristics; for instance, unique cell wall structures [35]. A recent report suggests that rare prokaryotes play a crucial buffering role in biotic community membership and stability, besides acting as a genetic reservoir in the face of environmental perturbation [36]. It is well understood that the marine microbial community plays an enormous role in recycling global nutrients [37], and most currently established biological pathways are built on well-known prokaryotes. However, it is less understood that certain underexplored prokaryotes also have biological roles in the environment, in particular in geochemical cycles of carbon, nitrogen, and sulfur [34,38,39]. In the past, researchers only examined the effects of abiotic factors on overall microbial diversity, particularly the predominant taxa [40][41][42][43]. Researchers have started to investigate how environmental factors affect the diversity of the rare prokaryotes instead of the major microorganisms [44]. As we broaden horizons of underexplored prokaryotes, genomic evolution becomes better understood, and missing links in gene transfer are discovered, so that revising the tree of life may be possible.
The second primary motivation of working in the field of underexplored bacteria is their association with health and diseases. Researchers suspect that certain underexplored bacteria are related to infections in humans, animals, or plants. For example, the candidate (or candidatus in Latin) Liberibacter asiaticus is pathogenic to potato, carrot, tomato, and citrus [45]. The threat from these bacteria is so severe that the US Environmental Protection Agency (EPA) is currently in the process of allowing citrus growers to spray antibiotics to control pathogenic Liberibacter species [46]. In humans, the growth of an underexplored prokaryote, namely candidate Borkfalki ceftriaxensis, was observed following antibiotic treatment with ceftriaxone.
Additionally, it is now understood that other than viruses, bacteriophages, fungi, and bacteria, the candidate phyla radiation (CPR) superphylum group of bacteria interact with each other in the human oral environment and collectively may impact human health [34]. On the other hand, the search for new antibacterials also needs to continue. It needs to be determined whether underexplored microorganisms are important sources of new antibacterial agents. The US Government Defense Advanced Research Projects Agency (DARPA) conceptualized the Pathogen Predators Research Program with an enacted budget of 3.4 billion USD for 2019. One of DARPA's projects was searching for predatory bacteria against pathogens [47]. To date, Bdellovibrio bacteriovorus (from a genus consisting of four type strains) and Micavibrio aeruginovorus (from a monotypic genus) are the only predatory bacteria known to prey upon >100 different human pathogens [48][49][50].
The third primary motivation is the new applications-Blue Ocean Strategy. Protein sequences of underexplored prokaryotes exhibit low similarities to other well-established proteins. The exploration of underexplored prokaryotes and their macromolecules and natural products is regarded as the 'blue ocean strategy' in science (a term coined from a book of the same title by Mauborgne and Kim [51]). The author defined the 'blue ocean strategy' as "the simultaneous pursuit of differentiation to open up a new market space and create new demand". Therefore, discovering new strains or native and novel macromolecules may establish new, if not, better applications. In some reports, the total number of genes annotated as hypothetical proteins represent half of the total number of genes [52]. Some archaeal genomes had up to 80% hypothetical proteins [53]. It would be interesting to identify new potential applications of these unknown genes. In addition, exploring rare prokaryotes is an opportunity for an offensive patent strategy of potentially lucrative novel bioeconomy.

Why Are Great Proportions of Prokaryotes Unculturable?
Researchers blame the prokaryotes since much about them remains 'black box information' or non-reproducible under the designated laboratory setup. For example, (i) cells may need unknown special culture requirements: uncommon nutrients, a narrow temperature or pH range, or unusual compounds to support their growth. (ii) Cells from the environment are in the dormancy stage, and none of the resuscitation approaches can make them appear on the Petri dish, and (iii) having a smaller genome size causes most candidate CPR to lack numerous biosynthesis pathways, to be absent of ATP synthase, and to lack the electron transport chain complex [54]. As a result, Dombrowski et al. [55] concluded that CPR and DPANN members are unable to grow individually but, rather, rely on resources contributed by neighboring bacteria through symbiosis, parasitism, or other relationships. Despite such biological interactions also existing in culturable microorganisms [56], CPR and DPANN members are relatively more delicate.

How Should Underexplored Prokaryotes Be Cultured?
It is not impossible to isolate underexplored prokaryotes from the environment. From the authors' analysis using data from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), at least 1000 genera are single-species representatives (monotypic or monospecies) which were once categorized as 'unculturable' (Table 4). Although some monotypic strains require stringent isolation or cultivation strategies, many of the bacteria are not at all demanding to cultivate. Although the suggestions provided hereafter are not new, however, microbiologists in their early career may be unaware of them.
The first suggestion is paying attention to isolation and growth preparation. Autoclaving agar with phosphate remarkably lowers the total number of colonies that grow on the agar plates [57]. Agar contains some inhibitors that prevent the growth of some prokaryotes; gelrite is a better solidification agent. The enrichment of any environmental sample (e.g., soil or sediment) using atypical chemicals may enhance the probability of isolating rare prokaryotes. For example, Aanderud [58] used heavy water (H 2 18 O) to rewet soil samples. Interestingly, after introduction of heavy water, the abundance of initially recognized rare microorganisms increased considerably, from hardly detectable to the dominant proportion of the community. The second suggestion is: reconsider sampling sites. In order to increase the success rate of discovering undomesticated species, individuals who conduct the sampling should perhaps consider using some hard-to-reach locations where human activities are at a minimum. For example, hydrothermal vents in ocean basins, boiling geothermal springs, and caves, such as those at Mount Roraima in South America; rainforests, such as the deep Amazon or Malaysia's Kinabalu National Park; parched dry places, such as the Atacama Desert colourful lakes, such as those of Indonesia's Kelimutu volcano; or deep-cold places, such as Antarctica (below 2.75 km in depth). Members of the Extreme Microbiome Project are interested in poly-extremophiles, for example, thermophiles in the Hell gas crater of Turkmenistan, and halophiles in Lake Hiller of Australia [59]. Readers are referred to a recent review article that discusses current knowledge on extremophiles [60].
The third suggestion is modifying cultivation strategy. (i) In situ cultivation: If possible, researchers could consider using a device such as isolation chips (ichips) to maximize the number of individual colonies [61,62]. The idea behind the ichip is to dilute the environmental samples using molten agar, trapping each cell into a hollow compartment sealed with a membrane, and returning the ichip to the environment (Figure 2). Using the ichip, a new class of antibiotics (teixobactin) was discovered from an unculturable bacterium, Eleftheria terrae [47]. (ii) Ex situ cultivation: Recently, Chaudhary et al. [63] described a new bacterial cultivation device termed as a diffusion bioreactor (Figure 2). Using this approach, the authors isolated 35 previously uncultured strains from phyla Proteobacteria, Firmicutes, Actinobacteria, and Bacteroidetes. The diffusion bioreactor is more flexible than that of the in situ ichip method. As the diffusion bioreactor is performed in the laboratory, researchers have a better selection of abiotic or experimental parameters, for instance, pH, temperature, moisture, nutrients, and timing. Lately, Sun et al. [64] discussed several strategies to improve archaeal cultivation.

Exploring Unculturable Prokaryotes Using Metagenome-Assembled Genomes (MAG)
Even with intensive isolation efforts, a certain proportion of prokaryotes escape cultivation-for example, candidate phyla, unculturable superphylum, such as the bacterial CPR and archaeal DPANN, and other unexplored or unknown phyla (Figure 1). The exact number of unexplored phyla is unknown. Examples of a few candidate phyla are listed in Table 1, and this number is expected to rise. The naming of uncultured taxa could be confusing due to inconsistencies in nomenclature [92], dynamic updates and renaming, and differences of information in major databases, for instance, those of the NCBI Taxonomy, Silva, and Genome Taxonomy Database (Table 1) [2,8].
To date, unculturable prokaryotes bacterial CPR membership has been conferred to more than 70 Candidate phyla [92]. There are many putative archaea phyla identified via culture-independent approaches [53,64]. The cumulative 16S rRNA amplicon sequencing and genomic data have led to the proposal of several archaea superphyla. For instance, DPANN is the largest archaea superphyla with >24,000 deposited 16S rRNA sequences listed in NCBI taxonomy and Silva databases (September 2019). The readers can refer to a recently published review article on DPANN (candidate Aenigmarchaeota, Altiarchaeota, Diapherotrites, Hadesarchaeaeota, Huberarchaeota, Micrarchaeota, Nanohaloarchaeota, Pacearchaeota, Parvarchaeota, and Woesearchaeota) for more insights on the genomic features, lifestyle, and evolution [55]. Other unculturable archaea superphyla include Asgard (also known as Asgardaeota) and TACK (also known as Proteoarchaeota) [64]. The TACK superphyla group consisted of Nitrosopumilales, Nitrosotalea, Nitrosophaerales, Nitrospcaldales, Geothermarchaeota, Aigarchaeota, Bathyarchaeota, Marsarchaeota, Nezhaarchaeota, Geoarchaeota, and Verstraetearchaeota. However, the TACK is undergoing dynamic membership updates and renaming, and the taxonomy affiliation of certain members are uncertain.
MAG is one of the most effective ways to glimpse the genomes of unculturable prokaryotes. Anantharaman et al. [33] described 47 newly discovered phyla using 2540 reconstructed MAGs and elucidated that microbiome biological pathways are cross-linked. Danczak et al. [93] recently discussed the putative functions of 32 already known CPR Candidate phyla in carbon processing and nitrogen cycling. An impressive large-scale reconstruction of 7903 MAGs was performed, which provided the first genomic representatives of new rare candidate bacteria and archaea [17]. Accordingly, all the assembled genomes have at least 50% completeness, and nearly half of the total MAGs are ≥90% complete with less than 5% contamination. In a separate study, MAG analyses of two archaea from a hyperthermal hot spring identified unrecognized methane-metabolising sequences outside of the phylum Euryarchaeota [94]. In another work, Kadnikov et al. [95] recovered a complete genome of the candidate phylum BRC1 using MAG analysis of a deep subsurface thermal aquifer. For more information, readers can refer to a recent review article by Quince et al. [96] that covers experimental design, sampling, and analysis using shotgun metagenomics for MAG. A summary of MAG-related research and bioinformatics tools is given in Table 5. CheckM is a tool used to determine genome completeness and identify contaminant sequences [97].

Potential Applications of Culturable Rare Prokaryotes
A recent study has shown that depths of 6−11 km below sea level are heavily polluted by microplastic (fragment size ≤ 5 mm) [113]. A recent review article summarized current milestones in the use of microbial enzymes for modifying or degrading different categories of plastics, for instance, polyurethane and polyethylene terephthalate (PET) [114]. Unfortunately, the degradation process of highly crystallized and packed plastics is still considered slow and ineffective. Readers are advised to read a review on marine microbial adaptation, interaction, and degradation of microplastics [115]. The biodegradation of recalcitrant plastics involves various groups of enzymes; including cutinases, laccases, manganese peroxidases, lignin peroxidases, alkane hydroxylases, tannases, hydroquinone peroxidases, ureases, esterases, lipases, proteases, and polyester hydrolases, although each with a different extent of degradations [114]. Enzymes from various genera have been described to have some form of plastic degrading ability. Delftia, Thermobifida, and Ideonella being among the genera with limited type strains [114]. Ideonella sakaiensis is able to use PET as carbon source [116]. Enzymes from I. sakaiensis and Thermobifida fusca have been extensively examined with regards to the aspects of enzyme catalytic mechanism, structure, and function [117][118][119][120]. In nature, PETases (PET hydrolase) are esterase active on ester bonds [116,120]. In a separate work, Danso et al. (2018) mined 853 gene sequences using a metagenomics approach and showed that PETases are mainly distributed in culturable Actinobacteria, Proteobacteria, and Bacteroidetes [121], and among the sources, Caldimonas and Methylibrium can be further explored. Additionally, PETase homologues are also found in other groups of bacteria [118].
Efforts have been invested in harnessing rare marine Actinobacteria. Examples of underexplored Actinobacteria are Actinoalloteichus, Salinispora, Marinactinospora, and Actinosynnema. These prokaryotes produce active compounds with cytotoxic, antibacterial, antifungal, and antimalarial activity [122]. The phylum Actinobacteria is one of the largest phyla in terms of total isolates and sequences deposited in databases (Figure 1, Table 1). However, the isolation and discovery of new Actinobacteria species are continuing. In a large-scale screening, Idris et al. [123] elucidated that 16% of the total sequence reads associated to this phylum could only be assigned up to class level based on the EzTaxon-e database, and these new taxa had relatively low similarity to already-described Actinobacteria genera.
Lambrechts and Tahon [124] recently summarized the progress of Antarctic microbiology studies and listed many monotypic genera of rare bacteria. More attention should be given to psychrophilic microorganisms, as their enzymes are of potential use in biotechnology, particularly for applications that require lower temperatures [125]. There are many underexplored psychrophilic prokaryotes; for example, a monotypic Raineyella antarctica was recently isolated and sequenced [126]. Sheridan et al. [127] described Rhodoglobus 16 years ago while Li et al. [128] first reported monotypic Marisediminicola antarctica ca. 10 years ago. Unfortunately, neither genera have received much attention. Few researchers are interested in studying psychrophiles, possibly because many of these bacteria grow slowly and most of their proteins (enzymes) are sensitive to higher temperatures.
Industrial enzymes, such as hydrolases for starch, cellulose, hemicellulose, lipids, and esters, are biocatalysts that have been extensively targeted using culture-based approaches. Many rare bacteria type strains harbour interesting genes, and a few are listed here. Melioribacter roseus is a facultative anaerobe and a thermophile that is rich in various glycosyl hydrolase (GH) genes [129][130][131]. The bacteria Siansivirga zeaxanthinifaciens, a mesophile, and Alkalitalea saponilacus, an anaerobe and alkalophile, also harbour a broad range of GHs [132][133][134][135]. All the bacteria mentioned above are monotypic of their respective genus. Jeotgalibacillus is another example of an underexplored genus whose enzymes may be of use [136]. Liew et al. [137] recently elucidated that J. malaysiensis produces glucose-tolerant β-glucosidase.
Rhodothermaceae (thermophilic), Rubricoccaceae (mesophilic), Salisaetaceae (halophilic), and Salinibacteraceae (halophilic) are families of the order Rhodothermales. To date, this order consists of ten genera, each with no more than three species. The order is small, having only 16 validly described type strains. Rhodothermus spp. produce important cellulosic and hemicellulosic hydrolases [138,139]. However, none of the other genera in the order Rhodothermales have been studied for counterpart enzymes. Lately, Park et al. [140] isolated Roseithermus sacchariphilus gen. nov., sp. nov. (strain MEBiC09517 T ) from sediment collected from a coastal area. Due to the high genome-to-genome similarity, our isolated bacterium Rhodothermaceae RA is the subspecies of strain MEBiC09517 T [141,142]. The bacterium R. sacchariphilus RA has 57 glycosyl hydrolase sequences. We have cloned and characterized two novel xylanase enzymes [143,144]. For one of the novel enzymes that we characterized, the protein sequence showed less than 50% sequence identity to well-characterized counterpart enzymes. The authors believe that all members of the order Rhodothermales should be given more consideration for the discovery of enzymes.
The International Society of Rare Sugar (http://www.isrs.kagawa-u.ac.jp) has defined rare sugars as monosaccharides and their derivatives that rarely exist in nature [145]. These fine chemicals (i.e., D-allose, D-psicose, D-gulose, D-sorbose, and L-ribose) are valuable and have numerous applications in food products and sweeteners, pharmaceuticals, and agriculture [145][146][147]. The conversion of natural sugars to rare sugars involves several enzymatic reactions involving isomerases, epimerases, and oxidoreductases [145]. The majority of enzymes already applied in the industry have been sourced from cultured bacteria. To the best of our knowledge, attempts of finding underexplored culturable prokaryotes as well as gene-mining using a metagenomic approach to produce rare sugar are scarce [147].
Many monotypic bacteria remain to be discovered and could provide insights into fundamental sciences and offer possible biotechnological applications. Table 4 summarizes some monotypic bacteria with genome information. Other records of monotypic bacteria and genera are available on the LPSN database and Wikipedia (http://www.wikipedia.org; keyword: monotypic bacteria. Note: the list may not be auto updated). Readers who wish to have access to a periodically updated list of prokaryotes can refer to Prokaryotic Nomenclature on the official webpage of DSMZ.

Potential Research Topics in Unculturable Prokaryotes
For culturable rare prokaryotes, we store the cells on Petri dishes or freeze them as stock cultures. This strategy is different from exploiting resources from unculturable prokaryotes since we do not have cells of unculturable prokaryotes. Researchers only retain extensive digital information (contigs or scaffolds, annotated genes, or sequences of proteins) and a limited volume of eDNA. The easiest way to explore biological resources (genes and proteins) mined in MAGs or SAGs is by synthesizing the complete genes using commercial service providers, cloning the genes in a suitable vector, expressing the genes as recombinant proteins, examine the biochemical functions, or analyze the protein structures by methods such as X-ray crystallization. Although synthesizing several genes is affordable for most laboratories, this workflow is not flawless as the selection of the genes is pivotal. It is not always possible to pinpoint the right targeted genes due to knowledge limitations, or the expressed recombinant proteins are neither active nor function as anticipated.
Ocean and coastal bodies are the most significant reservoirs of diverse microbes and important sources for the discovery of prokaryotes. Tara Oceans is a well-known global ocean expedition involving hundreds of researchers. MAG information from this project is publicly available [107]. Other MAG data obtained from various environments are also deposited in public databases, such as the NCBI. By exploring putative genes of interest generated in-house or from sources such as the Tara Oceans expedition, researchers will have ample genes that encode biocatalysts or macromolecules that may be applicable in biotechnology. Examples of metagenome-derived biocatalysts and proteins, both from the common and underexplored prokaryotes, are listed hereafter; haloalkane dehalogenase, esterase, β-glucanase, keratinase, exonuclease, and endoglucanase [148][149][150][151][152][153]. Other examples of recent work using MAGs to mine biocatalysts and proteins are listed in [150,154]. Readers are encouraged to read some excellent review articles on the topic of MAGs [155,156].

Limitations and Future Directions of Prokaryote Discovery
The authors agree with the opinion of Garza and Dutilh [157] that scientists have only scratched the surface of the vast microbial world. The total number of phyla in the prokaryotic domain remains unknown (Figure 1). This review does not intend to exaggerate the importance of underexplored prokaryotes. We acknowledge that prokaryotes such as Bacillus are still one of the best workhorses for industrial applications. One needs to understand that there are many obstacles and limitations to exploring rare prokaryotes, despite the rewards it may offer, and a few will be discussed here.
Using single-cell genome amplification to explore unculturable prokaryotes is not straightforward. Hedlund et al. [158] reported that SAG is not an ideal approach if the environmental sample is too complicated. However, there have been a few successful applications. In most cases, genome assembles derived from SAG is shorter than MAG, depending on the type of samples [18,21,159]. On the other hand, SAG yields more focused reads than the wider MAG. As such, methods integrating SAG and MAG, such as the JGI's innovative microfluidic-based mini-metagenomic method, offers a more accurate complete genome [22]. However, not all laboratories have the resources to use microfluidic technology.
The invention of the ichip is interesting and nothing short of ingenious. Furthermore, all laboratories can convert affordable plastic pipette tip racks into a homemade ichip. According to the descriptions in Berdy et al. [160], it is possible to obtain a single type of cell or combination of a few strains in each chamber. We speculate that pairing the ichip with MAG would be much easier and cheaper than SAG-MAG methods ( Figure 2). However, this concept has not been experimentally proven.
Prokaryote genomes vary in size; for example, candidatus Carsonella ruddii has a tiny size of 0.17 Mb, while the Sorangium cellulosum strain So0157-2 has a very large size of 14.78 Mb [161,162]. The mean genome size for prokaryotes is~3.7 Mb (https://www.ezbiocloud.net) [163]. Based on statistical analysis, many Actinomycetes strains that produce important secondary metabolites have large genomes, approximately >8 Mb [122]. According to the article, the coding capacity increases with genome size [122]. It is not uncommon that prokaryotes with genomes smaller than 8 Mb also exhibit a complete set of secondary metabolite gene clusters as well as other important industrially-applicable proteins. However, unculturable CPR and the DPANN superphyla have very small genomes because they have minimal biosynthetic capacities [52,54,55]. Therefore, some unculturable prokaryotes, especially those with extraordinarily small genomes, may encode lesser industrially applicable proteins, thus, limiting their exploitation in biotechnology. The genus Bacillus is a favorite source of industrial enzymes because it has a sufficiently large genome and, thus, produces a broad range of proteins, compared to its counterpart Anoxybacillus, whose genome is approximately 50% shorter. Some of the commonly found genes, for instance enzyme CGTase, is present in Bacillus and Geobacillus, but absent in taxonomically-related Anoxybacillus due to genome shrinking [164,165]. Nevertheless, this does not mean that prokaryotes with a small genome have no potential use. For example, Caldicellulosiruptor saccharolyticus is an excellent thermophilic candidate for H 2 production and plant biomass degradation, despite its genome (2.9 Mb) being smaller than the mean size of prokaryote genomes [166]. Many hyperthermophilic archaea in general have smaller genomes, for instance, the genome size for Pyrococcus spp. is~1.9 Mb [167]. DNA polymerase from Pyrococcus is one of the high-fidelity enzymes that has been long commercialized by New England Biolabs.
We want to emphasize that MAG is not a magic staff. MAG is still not a common approach that all laboratories can perform. Novel sequences can be found from MAGs, even when a pure culture is not available. It is possible that the assembly of reads is imperfect, and this can create an artefact gene. The translated recombinant protein can malfunction due to a mutation at important positions due to the artefact in the gene. Unfortunately, even with properly assembled genes, for instance, those identified in high-quality WGS, various sophisticated persistent problems remain unsolved; for example, failure due to protein misfolding in the expression hosts. In addition, determining the roles of hypothetical proteins (assembled in WGS, MAG, or SAG) is relatively more costly and time-consuming. Deducing the protein's function may require additional work, such as transcriptomic and in silico functional prediction. The crystallization of hypothetical proteins to identify clues from the protein structure is another potential method. Readers can refer to recent work on the discovery of hypothetical proteins [168][169][170][171]. The numbers of hypothetical proteins will expand rapidly as more underexplored genomes are sequenced; however, the speed to understand proteins' functions remain sluggish [53].

Conclusions
We believe that all creatures, including underexplored prokaryotes, have the potential for as-yet-unknown purposes. The cultivation of underexplored prokaryotes is somewhat challenging; yet, with the right experimental approach, a proportion of these cells can be isolated and grown in laboratories. Researchers can consider modifying their standard laboratory practices and try some different methods for selecting sampling sites, resuscitation approaches, enrichment techniques, or explore the state-of-the-art cultivation methods, for instance in situ and ex situ cultivation. Different DNA sequencing strategies, i.e., WGS, SAG, and MAG, allow researchers to glimpse the genomic contents of underexplored prokaryotes. Overall, it appears that the MAG approach is a more reliable strategy for discovering unculturable microbes. If assembly is performed carefully, MAG methods can be used to generate accurate genome information from uncultivated bacteria as well as underexplored culturable bacteria and will eventually become a standard tool. However, functional-based metagenomics remains expensive and is not affordable on a larger scale. The quality of MAGs is highly dependent on the type and preparation of samples, and MAG requires computational capacity for processing large quantities of sequencing read data and downstream analyses. It is believed that exploring rare prokaryotes is interesting, yet requires due diligence and wise decision-making to overcome all possible challenges.